During orthotopic liver transplantation (OLT), hyperkalemia frequently follows massive transfusion, renal insufficiency, or a sudden release of potassium from the graft on reperfusion.1,2 We describe a case of acute hyperkalemia that was provoked by acute hepatic necrosis during the pre-anhepatic stage in living donor liver transplantation (LDLT). Despite initial treatment, including calcium insulin and glucose, sodium bicarbonate, and furosemide,3 hyperkalemia did not resolve until nebulized salbutamol was administered. The efficacy of a selective β2-agonist, such as salbutamol, in lowering serum potassium has been established in normal volunteers4 and in chronic renal failure patients.5–9 However, there are no data regarding its potential efficacy as a treatment option for hyperkalemia as a result of OLT, especially during the anhepatic stage. We report the successful use of nebulized salbutamol in combination with conventional treatments to prevent a life-threatening increase in serum potassium during the pre-anhepatic stage in LDLT. The patient provided written consent for publication of this report.

Case report

A 50-yr-old man (177 cm, 83 kg) suffering from end-stage hepatitis B liver cirrhosis was admitted for LDLT. He had a history of four transarterial chemoembolizations and an ethanol injection for the recurrence of hepatocellular carcinoma. There was no evidence of hepatorenal syndrome and no history of diabetic mellitus or renal dysfunction. His serum potassium concentration was normal (4.4 mEq l−1) one day before surgery, as were his blood urea nitrogen (BUN; 20.5 mg dl−1) and creatinine (1.3 mg dl−1) levels. The model for end-stage liver disease score at OLT was 19. He had not taken any medication known to cause hyperkalemia within the previous two weeks.

Anesthesia was induced with fentanyl 100 μg i.v., thiopental sodium 250 mg i.v., and vecuronium 8 mg i.v., and was maintained with oxygen, air, and desflurane. Neuromuscular block was maintained with atracurium. Hemodynamic monitoring included right radial and femoral artery, central venous, pulmonary artery, and capillary wedge pressures, as well as cardiac output measurements. The surgery proceeded by way of the piggyback technique, with preservation of caval flow. Blood samples were taken from a catheter in the right radial artery. The initial serum potassium concentration was 4.6 mEq l−1 (Table 1). At the mid-point of the hepatectomy, the concentration increased to 5.9 mEq l−1 (Table 1) and then further increased to 6.6 mEq l−1 (Table 1), despite the administration of calcium gluconate (1 g i.v.) and an infusion of sodium bicarbonate (40 mEq) and dextrose (50%, 50 ml) with 10 U of regular insulin. At this stage, the electrocardiogram (ECG) monitor showed tall T waves, absent P waves, and a normal QRS complex. Although hourly urine output remained 80–100 ml h−1, furosemide 10 mg i.v. was administered. Subsequently, the administration of dextrose (50%, 50 ml i.v.) with 10 U of regular insulin was repeated. Approximately 2 h after the initial anti-hyperkalemic interventions, hyperkalemia persisted. At this point, neither potassium-containing fluids nor packed red blood cells had been administered, nor was there evidence of significant metabolic acidosis (Table 1). The cause of the hyperkalemia was unclear, but it was thought to have been related to the surgical procedure. Due to anatomic factors (a caudate lobe wrapping around the IVC), about 90 min had elapsed from the division of the hepatic artery and the portal vein to the clamping of the suprahepatic inferior vena cava (IVC), and it was presumed that the persistent hyperkalemia resulted from loss of potassium from ischemic liver cells into the systemic circulation.

Table 1 Arterial blood gases, and potassium, ionized calcium, and glucose concentrations during liver transplantation
Full size table

We decided to incorporate nebulized β2-agonist therapy into the combined regimen of insulin, glucose, and sodium bicarbonate, and we determined that hemodialysis would have to be considered if hyperkalemia did not resolve after one more therapeutic cycle of this combined therapy. Thus, we changed the anesthetic machine to one with a different ventilator (Savina, Drager Medical, Lübeck, Schleswig-Holstein, Germany), which enabled the nebulized treatment to be administered during positive pressure ventilation. Nebulized salbutamol (10 mg in 3 ml of normal saline) was administered over 20 min. During this period, the desflurane anesthesia was supplemented with intermittent bolus injections of propofol and fentanyl, and the infusion of dextrose (50%, 50 ml) with 10 U of insulin and sodium bicarbonate (40 mEq) was continued. Except for transient tachycardia at the initiation of the salbutamol treatment, no complication was observed. As there was little time at this point to reperfuse the liver graft, we simultaneously mobilized the necessary personnel and equipment for possible hemodialysis. However, within 20 min of receiving the combined therapy (10 min before reperfusion), serum potassium had decreased to 4.7 mEq l−1 (Table 1). The liver graft was vigorously flushed with 5% albumin (250 ml) via the portal vein before the reperfusion. Five minutes after the administration of calcium gluconate 1 g i.v., the liver graft was reperfused, and postreperfusion syndrome did not develop.

Two hours after the initiation of salbutamol treatment, serum potassium gradually increased to 5.8 mEq l−1 (Table 1), and a gradual peaking of the T wave was observed again on the ECG monitor. The patient was treated once again with insulin and nebulized salbutamol 10 mg i.v., and serum potassium again normalized to 4.6 mEq l−1 (Table 1). Further surgery was completed uneventfully, and the serum potassium concentration remained within the normal range. The operation lasted 12 hr15 min, and the anhepatic stage lasted 110 min. The estimated blood loss was 1800 ml; 2 U of packed red blood cells were administered during the surgery.

Discussion

Although hyperkalemic episodes are most common and important immediately after graft reperfusion, they can also develop at the pre-anhepatic stage during OLT.2,10 Xia et al.2 found that 10.2% of the patients develop hyperkalemia (≥5.5 mEq l−1) in the pre-reperfusion period and that an increased baseline potassium concentration and red cell transfusion were independent predictors of hyperkalemia in this period. The preoperative or intraoperative administration of several medications, including angiotensin converting enzyme inhibitors, spironolactone, trimethoprim, succinylcholine, heparin, and epsilon-aminocaproic acid has also been identified as a risk factor for intraoperative hyperkalemia.11,12 However, in the present case, the patient had not received any of these medications or a blood transfusion before the development of hyperkalemia, and his preoperative serum potassium concentration was normal. Furthermore, the development of hyperkalemia cannot be explained by either an excessive input of potassium or a primary failure of renal potassium excretion.

When considering the events of this case and the time course of the hyperkalemia (during the late hepatic dissection period), extensive cellular potassium loss as a result of surgical manipulation was the most likely cause. At the critical point of the hyperkalemic response, the hepatic artery and portal vein had already been divided, but the clamping of the suprahepatic IVC had been delayed for about 90 min because the caudate lobe was firmly wrapped around the IVC. Therefore, it is likely that progressive ischemic hepatic necrosis ensued during the interruption of hepatic flow and that a quantity of potassium was released from the necrotic liver cells and entered the systemic circulation. Ischemic liver cells can release large amounts of intracellular potassium,13 and the potassium concentration in the systemic circulation is known to increase when the hepatic circulation is interrupted for periods longer than 20 min.14,15 Moreover, hyperkalemia is more marked during longer periods of hepatic ischemia16 and can be sustained for 2 h, even when circulation is restored.15,16

A similar episode of severe hyperkalemia has been reported in the setting of LDLT, in which the piggyback technique was used.10 The piggyback technique, with preservation of the native IVC and direct anastomosis between the donor’s IVC and the recipient’s hepatic vein, avoids retrocaval and adrenal dissection, thereby leading to reduced bleeding and a shortened anhepatic stage. Furthermore, it offers the advantages of a low incidence of renal impairment and improved hemodynamic stability during the anhepatic stage without the use of venovenous bypass.17 However, anatomical challenges, such as intensive inflammatory adhesions or a caudate lobe wrapped tightly around the IVC, often limit its use.17 In such cases, the anesthesiologist should remain vigilant to prevent hepatic flow interruption longer than 20–30 min and to monitor closely for the possible development of hyperkalemia.

For the management of acute intraoperative hyperkalemia, regardless of the underlying cause, three main approaches have been recommended: 1) to oppose the direct toxic effects of hyperkalemia on cell membranes by infusing calcium salts; 2) to promote cellular uptake of potassium using insulin with glucose, a β2-agonist, or possibly sodium bicarbonate, either alone or in combination; and 3) to remove potassium from the body using diuretics or hemodialysis.3,12

In this case, we administered relatively small doses of sodium bicarbonate and furosemide (80 mEq and 10 mg, respectively) for intractable hyperkalemia, because the patient had satisfactory urine output and no documented metabolic acidosis. The potassium-lowering effects of insulin have been demonstrated in the setting of OLT.18 Insulin effectively decreases the serum potassium concentration within 15 min, and its effect lasts for at least 60 min, even in the anhepatic phase. However, the initial treatment, which included bicarbonate, insulin, and furosemide, failed to normalize the serum potassium concentration. During the initial uncontrolled hyperkalemic episode, the potassium influx into the systemic circulation may have exceeded the patient’s renal excreting ability and the transcellular shifting capacities of the administered insulin (initial: 20 U, total: 40 U) and sodium bicarbonate. Ultimately, persistent hyperkalemia resolved after administering nebulized salbutamol. We believe that the patient’s hyperkalemia was relatively refractory to the conventional treatments, but resolved only after the introduction of multimodal therapy incorporating the effects of nebulized salbutamol with initial therapy.

Generally, treatment with nebulized salbutamol for hyperkalemia requires a dose of 10–20 mg delivered over 10–30 min.3,5 The onset of action is usually within 15–30 min, with a peak effect between 30 and 60 min, and its effect lasts for 4–6 h.3,5,12 The onset time and the duration of its hypokalemic effect are similar to those of insulin and sodium bicarbonate.12 In this case, we administered a total of 20 g of nebulized salbutamol over two cycles. The hypokalemic response is dependent on the dose administered, as greater efficacy was reported in patients receiving 20 mg vs. 10 mg of nebulized salbutamol.9

Salbutamol stimulates β2-adrenergic receptors in muscles and the liver, and then it enhances cellular uptake of potassium with direct activation of Na+–K+ ATPase via a rise in the cyclic AMP system; its effect is independent of insulin. Although the anhepatic stage of OLT excludes any hepatic function, in the present case, nebulized salbutamol reliably reduced the serum potassium concentration even in the anhepatic stage, possibly by an extrahepatic transcellular potassium shift. Moreover, insulin and β2-agonists are known to exert additive hypokalemic effects through different intracellular signalling mechanisms.7 When sodium bicarbonate is combined with insulin or a β2-agonist, it also has a synergistic effect in lowering the serum potassium concentration.8,19 Previous studies found that 20 mg of nebulized salbutamol lowered serum potassium by 1.21 ± 0.19 mEq l−1 in tandem with the administration of insulin 10 U i.v., while nebulized salbutamol 15 mg lowered serum potassium by 0.96 mEq l−1 when given after sodium bicarbonate (2 mEq kg−1 i.v.).8,20 In our case, despite a sustained influx of intracellular potassium to the systemic circulation, our triple combined therapeutic regimen reduced serum potassium by 2 mEq l−1. Therefore, it is efficacious to use a regimen combining these three agents, particularly for acute, severe hyperkalemia.

A major concern regarding the use of salbutamol is a paradoxical elevation in serum potassium concentrations in the first minutes following nebulization.5,21 Although this response is minor (0.15 mEq l−1)5 or transient (less than 3 min),5,21 in cases of excessive hyperkalemia, it is prudent to administer calcium salts immediately before initiating nebulized salbutamol.

In summary, this case suggests that a delay in clamping the suprahepatic IVC for a significant period after the interruption of both the hepatic artery and the portal vein can cause the development of acute severe hyperkalemia in patients undergoing OLT using the piggyback technique. In such situations, careful intraoperative ECG monitoring with frequent measurements of serum potassium are important. During liver transplantation, treatment with nebulized salbutamol, insulin, and sodium bicarbonate is an effective therapeutic combination to treat hyperkalemia.